HELM AI Kernel Ecosystem

June 19, 2026 · View on GitHub

This map helps contributors find high-value integration and upstream contribution targets.

Integration Map

flowchart LR
    MCP["MCP registries and docs"] --> Kernel["HELM execution boundary"]
    Gateway["AI gateway projects"] --> Kernel
    Guardrails["Security scanners and guardrails"] --> Kernel
    Frameworks["Agent frameworks"] --> Kernel
    Kernel --> Proof["Receipts and EvidencePacks"]
    Proof --> Listings["Awesome lists and ecosystem docs"]

Lane	Adjacent projects	HELM contribution angle
MCP registries and docs	modelcontextprotocol/registry, modelcontextprotocol/docs, modelcontextprotocol/servers	Tool metadata, schema-pin, authz, and fail-closed examples.
Remote tool routing infrastructure	envoyproxy/ai-gateway, microsoft/mcp-gateway, Portkey-AI/gateway, BerriAI/litellm, openziti/llm-gateway	Policy payloads, remote tool authorization, schema fidelity, and receipt-oriented audit fixtures.
Developer code intelligence	DeusData/codebase-memory-mcp	Local read/query graph intelligence for HELM source review; side-effecting MCP tools still route through HELM governance.
Security scanners and guardrails	promptfoo/promptfoo, NVIDIA/garak, NVIDIA-NeMo/Guardrails, guardrails-ai/guardrails	Deny/escalate examples, reproducible security tests, and evidence outputs for agent tool execution.
Agent frameworks	openai/openai-agents-python, pydantic/pydantic-ai, agno-agi/agno, crewAIInc/crewAI	Minimal integration guides for routing proposed tool calls through HELM before dispatch.
Awesome lists	punkpeye/awesome-mcp-servers, appcypher/awesome-mcp-servers, wong2/awesome-mcp-servers, corca-ai/awesome-llm-security, ProjectRecon/awesome-ai-agents-security, e2b-dev/awesome-ai-agents	Submit concise listings after the quickstart, demo asset, and contributor backlog are visible.